TINA/Spice: Loop compensation for DC/DC Converter

The idea is that there should be no more than 1 effective pole (# poles - # zeros) at crossover, which would mean that the phase dip from DC would be between 90 and 180 degrees depending on the location of the poles and zeros before the crossover.

If you crossover at -40db/decade, you already have more than 180 degrees phase shift and the circuit is unstable (since it is a negative feedback system)

Nikhil, thankyou for quick respone.
If there is only one effective pole at the crossover frequency then how we can get the phase margin between 45 & 60 degree to make the loop stable?
As per my understanding if there is only one effective pole then at that particular frequency phase shift will be 90 degree. Kindly suggest.

Well, it really depends on what frequency the poles and zeros are located. For example, If there are are 2 poles close to the origin (DC) and 1 zero which is close to the crossover several KHz away, it will not be able to compensate the phase dip from the 2 poles before the crossover which will result in a phase dip still be close to 180 degrees. In theory, you would want the zero to be at least 1 decade away from the crossover to get ~45 degrees Phase margin in this case.

Thanks, got it.
In some application note, it's mentioned that the phase margin range should be between 45 & 60 degree, what happens if it goes beyond the 60 degree (means the zero is near to crossover than expected) or below 45?
I was trying to see the response by changing value of compensation components (on TPS54560) for DFMEA purpose, in one case the phase margin was a bit high then the 60 degree (still magnitude is crossing at -20db/ decade and phase crossover frequency is greater than the gain crossover), but I didn't see kind of oscillation at the output and ripple in the output is within limit.